As one of the most important families of drug targets known to date, G protein coupled receptors (GPCRs) continue to be the subject of considerable research efforts directed towards discovering improved therapeutics. Although the recent availability of several high-resolution crystal structures of GPCRs enables a higher success rate in the discovery of novel compounds targeting these receptors, the accurate prediction of the in vivo efficacy of these compounds is what is really required for the rational discovery of improved therapeutics. Although binding affinity, i.e., the strength of association of a drug to its recepto at equilibrium, has often been used as an appropriate surrogate for in vivo efficacy, retrospective assessments of successful drugs in the market suggest that kinetic quantities, such as those that regulate the lifetime of the drug target complex, may be as important as, or even more important than, binding affinity for the prediction of the efficacy and/or safety of a drug in a liing organism. The overall goal of this application is to delineate an efficient computational strategy that is capable of predicting accurate kinetic quantities related to GPCR ligand binding in addition to identifying correct receptor binding sites and ligand binding modes in a completely flexible receptor embedded into an explicit membrane model. This strategy is expected to complement current rational drug design approaches for GPCRs, thus allowing the identification of better candidates for clinical drug development.